Change logic to check tx_avail < nsegs + reserved_space
Details
Details
- Reviewers
kmacy scottl hiren gnn - Group Reviewers
Intel Networking - Commits
- rS298224: Correct possible underflow conditions when checking for available space
Diff Detail
Diff Detail
- Repository
- rS FreeBSD src repository - subversion
- Lint
Lint Passed - Unit
No Test Coverage - Build Status
Buildable 3221 Build 3254: arc lint + arc unit
Event Timeline
Comment Actions
Just tagging you guys to see if this makes sense. I saw the em(4) driver underflow this calculation with dtrace and suspected that there is a small chance the other Intel drivers may do the same in extreme conditions considering that the code is basically the same.